- 學位論文
基於分段式互補濾波器及耦合卡爾曼濾波器之實時同步定位與地圖構建
Real-Time 2D SLAM Based on Staged Complementary Filter and Coupled Kalman Filter
摘要
本論文完成在未知環境採用耦合卡爾曼濾波器的實時同步定位與地圖構建，耦合卡爾曼濾波器是基於擴增型卡爾曼濾波器的延伸型態，本論文的系統架構包含了九軸慣性感測器、平面測距雷達。研究目的是達成在未知環境中不斷更新地圖與估測載具姿態。此外，為了能更快速的從九軸管性感測器與便面測距雷達接收數據，將與Beaglebone Black連接的九軸慣性感測器與平面測距雷達在機器人操作系統架構中設定成基於筆記型電腦的MATLAB節點。在MATLAB系統中當互補濾波器透過加速度計、陀螺儀、電子羅盤估算角度的同時，耦合卡爾曼濾波器也在估測載具姿態、加速度誤差、牆壁特徵。而為了更好的估測加速度，在系統中會有一小段時間互補濾波器與耦合卡爾曼濾波器採分段式濾波進行合作，在實時系統中採用耦合卡爾曼濾波器，是因為比起擴增型卡爾曼濾波器，耦合卡爾曼濾波器在系統運算方面不會太多耗時。本論文將在最後討輪一些特殊環境問題。本論文所構建的系統已經在三種環境中完成時測，其中兩種是人造環境，一種是真實環境。
並列摘要
This research realizes real-time simultaneous localization and mapping(SLAM) in unknown environments with coupled Kalman filter(CKF) which is based on Extended Kalman filter(EKF). The system configuration comprises a vehicle which carries a 9-DOF IMU sensor, 2D scanning laser lidar. The objective is to update a map of an unknown environment and simultaneously estimate the vehicle's position. Moreover, to receive data faster, from a 9-DOF IMU sensor which is connected to a Beaglebone Black. Robot Operating System(ROS) is deployed and is set as a node of a MATLAB on a laptop. However, the lidar itself has a built-in ROS, and is also set as a node.In the MATLAB, a complementary filter is used to estimate angles by an accelerometer, a gyroscope, and a compass. CKF estimates the vehicle's position, accelerate's error, wall's feature at the same time. A complementary filter and CKF are staged in a short time. This thesis presents CKF which compares to EKF is a time-saving filter. Some spacial environments problems will also be discussed in this thesis. For experiments, three environments(two article and one real) have been tested.